Browse > Article
http://dx.doi.org/10.5389/KSAE.2022.64.1.039

Estimation of Optimal Training Period for the Deep-Learning LSTM Model to Forecast CMIP5-based Streamflow  

Chun, Beom-Seok (Department of Agricultural Civil Engineering, Kyungpook National University)
Lee, Tae-Hwa (Department of Agricultural Civil Engineering, Kyungpook National University)
Kim, Sang-Woo (Department of Agricultural Civil Engineering, Kyungpook National University)
Lim, Kyoung-Jae (Department of Rural Construction Engineering, Kangwon National University)
Jung, Young-Hun (Department of Advanced Science and Technology Convergence, Kyungpook National University)
Do, Jong-Won (Integrated Water Management Supporting Department, Rural Research Institute, Korea Rural Community Corporation)
Shin, Yong-Chul (Department of Agricultural Civil Engineering, Kyungpook National University)
Publication Information
Journal of The Korean Society of Agricultural Engineers / v.64, no.1, 2022 , pp. 39-50 More about this Journal
Abstract
In this study, we suggested the optimal training period for predicting the streamflow using the LSTM (Long Short-Term Memory) model based on the deep learning and CMIP5 (The fifth phase of the Couple Model Intercomparison Project) future climate scenarios. To validate the model performance of LSTM, the Jinan-gun (Seongsan-ri) site was selected in this study. We comfirmed that the LSTM-based streamflow was highly comparable to the measurements during the calibration (2000 to 2002/2014 to 2015) and validation (2003 to 2005/2016 to 2017) periods. Additionally, we compared the LSTM-based streamflow to the SWAT-based output during the calibration (2000~2015) and validation (2016~2019) periods. The results supported that the LSTM model also performed well in simulating streamflow during the long-term period, although small uncertainties exist. Then the SWAT-based daily streamflow was forecasted using the CMIP5 climate scenario forcing data in 2011~2100. We tested and determined the optimal training period for the LSTM model by comparing the LSTM-/SWAT-based streamflow with various scenarios. Note that the SWAT-based streamflow values were assumed as the observation because of no measurements in future (2011~2100). Our results showed that the LSTM-based streamflow was similar to the SWAT-based streamflow when the training data over the 30 years were used. These findings indicated that training periods more than 30 years were required to obtain LSTM-based reliable streamflow forecasts using climate change scenarios.
Keywords
Long Short-Term Memory (LSTM); CMIP5; streamflow; deep learning;
Citations & Related Records
Times Cited By KSCI : 4  (Citation Analysis)
연도 인용수 순위
1 Shin, Y. C., K. J. Lim, K. S. Kim, and J. D. Choi, 2007. Analysis of watershed runoff and sediment characteristics due to spatio-temporal change in land uses using SWAT model. KCID Journal 14(1): 50-56 (in Korean).
2 Cioffi, F., F. Conticello, U. Lall, L. Marotta, and V. Telesca, 2017. Large scale climate and rainfall seasonality in a Mediterranean Area: Insights from a non-homogeneous Markov model applied to the Agro-Pontino plain. Hydrological Processes 31(3): 668-686. doi:10.1002/hyp.11061.   DOI
3 Jimeno-Saez, P., J. Senent-Aparicio, J. Perez-Sanchez, and D. Pulido-Velazquez, 2018. A comparison of SWAT and ANN models for daily runoff simulation in different climatic zones of peninsular Spain. Water 10(2): 192. doi:10.3390/w10020192.   DOI
4 Kratzert, F., D. Klotz, M. Herrnegger, A. K. Sampson, S. Hochreiter, and G. S. Nearing, 2019. Toward improved predictions in ungauged basins: Exploiting the power of machine learning. Water Resources Research 55(12): 11344-11354. doi:10.1029/2019WR026065.   DOI
5 Li, W., A. Kiaghadi, and C. Dawson, 2021. High temporal resolution rainfall-runoff modeling using long-short-term-memory (LSTM) networks. Neural Computing and Applications 33: 1261-1278. doi:10.1007/s11430-010-0031-1.   DOI
6 Zhang, and K. F. Man, 1998. Time series prediction using RNN in multi-dimension embedding phase space. In Proc. 1998 IEEE International Conference on Systems, Man, and Cybernetics (Vol. 2, pp. 1868-1873). IEEE. doi:10.1109/ICSMC.1998.728168.   DOI
7 Abbaspour, K. C., 2013. Swat-cup 2012. SWAT calibration and uncertainty program-A user manual.
8 Arnold, J. G., P. M. Allen, and G. Bernhardt, 1993. A comprehensive surface-groundwater flow model. Journal of Hydrology 142(1-4): 47-69. doi:10.1016/0022-1694(93)90004-S.   DOI
9 Ayzel, G., L. Kurochkina, E. Kazakov, and S. Zhuravlev, 2020. Streamflow prediction in ungauged basins: benchmarking the efficiency of deep learning. In E3S Web of Conferences (Vol. 163, p. 01001). EDP Sciences. doi:10.1051/e3sconf/202016301001.   DOI
10 Boulmaiz, T., M. Guermoui, and H. Boutaghane, 2020. Impact of training data size on the LSTM performances for rainfall-runoff modeling. Modeling Earth Systems and Environment 6: 2153-2164. doi:10.1007/s40808-020-00830-w.   DOI
11 Bui, V., N. T. Le, V. H. Nguyen, J. Kim, and Y. M. Jang, 2021. Multi-behavior with bottleneck features LSTM for load forecasting in building energy management system. Electronics 10(9): 1026.   DOI
12 Dogan, S. Isik, T. Toluk, and M. Sandalci, 2007 Daily streamflow forecasting using artificial neural networks. In International Congress on River Basin Management (pp. 22-24).
13 Fu, M., T. Fan, Z. A. Ding, S. Q. Salih, N. Al-Ansari, and Z. M. Yaseen, 2020. Deep learning data-intelligence model based on adjusted forecasting window scale: Application in daily streamflow simulation. IEEE Access 8: 32632-32651. doi:10.1109/ACCESS.2020.2974406.   DOI
14 He, X., J. Luo, G. Zuo, and J. Xie, 2019. Daily runoff forecasting using a hybrid model based on variational mode decomposition and deep neural networks. Water Resources Management 33(4): 1571-1590. doi:10.1007/s11269-019-2183-x.   DOI
15 Hinton, G., N. Srivastava, and K. Swersky, 2012. Neural networks for machine learning lecture 6a overview of mini-batch gradient descent. Cited on, 14(8): 2.
16 Hochreiter, S., and J. Schmidhuber, 1997. Long short-term memory. Neural Computation 9(8): 1735-1780. doi:10.1162/neco.1997.9.8.1735.   DOI
17 Hu, C., Q. Wu, H. Li, S. Jian, N. Li, and Z. Lou, 2018. Deep learning with a long short-term memory networks approach for rainfall-runoff simulation. Water 10(11): 1543. doi:10.3390/w10111543.   DOI
18 Hussain, D., T. Hussain,, A. A. Khan,, S. A. A. Naqvi, and A. Jamil, 2020. A deep learning approach for hydrological time-series prediction: A case study of Gilgit river basin. Earth Science Informatics 13(3): 915-927. doi:10.1007/s12145-020-00477-2.   DOI
19 Kim, H. I., K. Y. Han, and J. Y. Lee, 2020. Prediction of Urban Flood Extent by LSTM Model and Logistic Regression. KSCE Journal of Civil and Environmental Engineering Research 40(3): 273-283 (in Korean). doi:10.12652/Ksce.2020.40.3.0273.   DOI
20 Husken, M., and P. Stagge, 2003. Recurrent neural networks for time series classification. Neurocomputing 50: 223-235. doi:10.1016/S0925-2312(01)00706-8.   DOI
21 Jung, S., H. Cho, J. Kim, and G. Lee, 2018. Prediction of water level in a tidal river using a deep-learning based LSTM model. Journal of Korea Water Resources Association 51(12): 1207-1216 (in Korean). doi:10.3741/JKWRA.2018. 51.12.1207.   DOI
22 Kannan, N., S. M. White, F. Worrall, and M. J. Whelan, 2007. Sensitivity analysis and identification of the best evapotranspiration and runoff options for hydrological modelling in SWAT-2000. Journal of Hydrology 332(3-4): 456-466. doi:10.1016/j.jhydrol.2006.08.001.   DOI
23 Maas, A. L., A. Y. Hannun, and A. Y. Ng, 2013. Rectifier nonlinearities improve neural network acoustic models. In Proc. icml (Vol. 30, No. 1, p. 3).
24 Mahsa, M., and T. Lee, 2018. Comparison of optimization algorithms in deep learning-based neural networks for hydrological forecasting: Case study of nam river daily runoff. Journal of the Korean Society of Hazard Mitigation 18(6): 377-384 (in Korean). doi:10.9798/KOSHAM.2018.18.6.377.   DOI
25 Pearson, K., 1895. Correlation coefficient. In Royal Society Proceedings (Vol. 58, p. 214).
26 Samantaray, S., and A. Sahoo, 2020. Appraisal of runoff through BPNN, RNN, and RBFN in Tentulikhunti watershed: A case study. In Frontiers in Intelligent Computing: Theory and Applications (pp. 258-267). Springer, Singapore. doi:10.1007/978-981-13-9920-6_26.   DOI
27 Le, X. H., H. V. Ho, and G. Lee, 2019. River streamflow prediction using a deep neural network: a case study on the Red River, Vietnam. Korean Journal of Agricultural Science 46(4): 843-856. doi:10.7744/kjoas.20190068.   DOI
28 Lee, G., S. Jung, and D. Lee, 2018. Comparison of physics-based and data-driven models for streamflow simulation of the Mekong river. Journal of Korea Water Resources Association 51(6): 503-514 (in Korean). doi:10.3741/JKWRA.2018.51.6.503.   DOI
29 Li, M., Z. Ma, and J. Du, 2010. Regional soil moisture simulation for Shaanxi Province using SWAT model validation and trend analysis. Science China Earth Sciences 53(4): 575-590. doi:10.1007/s11430-010-0031-1.   DOI
30 Liu, D., W. Jiang, L. Mu, and S. Wang, 2020. Streamflow prediction using deep learning neural network: Case study of Yangtze river. IEEE Access 8: 90069-90086. doi:10.1109/ACCESS.2020.2993874.   DOI
31 Gao, S., Y. Huang, S. Zhang, J. Han, G. Wang, M. Zhang, and Q. Lin, 2020. Short-term runoff prediction with GRU and LSTM networks without requiring time step optimization during sample generation. Journal of Hydrology 589: 125188. doi:10.1016/j.jhydrol.2020.125188.   DOI
32 Kratzert, F., D. Klotz, S. Hochreiter, and G. S. Nearing, 2021. A note on leveraging synergy in multiple meteorological data sets with deep learning for rainfall-runoff modeling. Hydrology and Earth System Sciences 25(5): 2685-2703. doi:10.5194/hess-25-2685-2021.   DOI
33 Xiang, Z., and I. Demir, 2020. Distributed long-term hourly streamflow predictions using deep learning-A case study for State of Iowa. Environmental Modelling & Software 131: 104761. doi:10.1016/j.envsoft.2020.104761.   DOI
34 Xiang, Z., J. Yan, and I. Demir, 2020. A rainfall-runoff model with LSTM-based sequence-to-sequence learning. Water Resources Research 56(1): e2019WR025326. doi:10.1029/2019WR025326.   DOI
35 Zhang, A., C. Zhang, G. Fu, B. Wang, Z. Bao, and H. Zheng 2012. Assessments of impacts of climate change and human activities on runoff with SWAT for the Huifa River Basin, Northeast China. Water Resources Management 26(8): 2199-2217. doi:10.1007/s11269-012-0010-8.   DOI
36 Zhihua, L. V., J. Zuo, and D. Rodriguez, 2020. Predicting of runoff using an optimized SWAT-ANN: A case study. Journal of Hydrology: Regional Studies 29: 100688. doi:10.1016/j.ejrh.2020.100688.   DOI
37 Shin, Y. C., M. H. Shin, W. K. Kim, K. J. Lim, and J. D. Choi, 2007. Estimation of streamflow discharges using Kajiyama equation and SWAT Model. KCID Journal 14(1): 41-49 (in Korean).
38 Tran, Q. K., and S. K. Song, 2017. Water level forecasting based on deep learning: A use case of Trinity River-Texas-The United States. Journal of KIISE 44(6): 607-612 (in Korean). doi:10.5626/JOK.2017.44.6.607.   DOI
39 Uniyal, B., J. Dietrich, C. Vasilakos, and O. Tzoraki, 2017. Evaluation of SWAT simulated soil moisture at catchment scale by field measurements and Landsat derived indices. Agricultural Water Management 193: 55-70. doi:10.1016/j.agwat.2017.08.002.   DOI